1

The question "IPv6 link-local addresses for virtual interfaces are identical - how to assign “link-unique”? " has been put on hold as offtopic on NE. Help page "What topics can I ask about here?" explicitly lists this as on-topic:

[...] On-topic: Questions about Unix / Linux Servers which are running Network Protocols or a Firewall

Questions about firewalling or network protocols such as OSPF running on linux / unix servers.

This is exactly what the question is about. It asks "What is the right way to obtain IPv6 addresses, which would be unique on the link, for such [VPN] interfaces?", pointing that this is needed for proper functioning of OSPFv3 on a Linux server.

How is this question not suitable for NE?

| |
4

Let me start by saying that I personally felt this question was in a grey area of being on/off topic, so I didn't vote to close.

However, as I pointed out in my comment, you are using invalid MAC addresses on your interfaces. So the correct way to get unique IPv6 addresses is to fix your configuration.

That configuration really has nothing to do with networking, IPv6 or OSPF. It is purely a operating system configuration. If you were to fix the OS problem, the networking issue would also be resolved.

| |
  • Thank you for the comment. But I am still looking for the answer to what I asked: "What is the right way to obtain IPv6 addresses...". For physical devices, MAC address are guaranteed to be unique by their vendor and means of their allocation. Is there anything similar for VPN links? If not, what is the best way/practice to configure the addresses? My belief is that this is very much on topic here. – Konstantin Shemyak Jul 16 '15 at 8:34
  • That was my thinking: The MAC address configuration is an off-topic server config issue. Can you change that, then just edit your Q with the results/additional info... the rest of your Q is clearly on-topic. – Craig Constantine Jul 16 '15 at 12:24

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .